Hi everyone,
in our solution in Kubernetes we have applications with an Ingress associated to a KongIngress.
All seemed to work fine, but yesterday for some test I deleted an application (deployment, service, ingress and kongingress) and I recreated it.
The field “Address” in Ingress was empty.
The ingress-kong pod was running, but logs for ingress-controller container stopped some days ago
and Last State was Error with Exit Code 255.
I resolved the issue restarting (delete) the ingress-kong pod, and automatically the Address field in Ingress was recovered.
How could I prevent it? Could I check other logs?
I0118 20:55:45.940971 1 controller.go:135] successfully synced configuration to Kong
I0118 20:55:49.821760 1 controller.go:135] successfully synced configuration to Kong
while this is the output of kubectl describe after my kubectl delete pod:
I’m not sure of the status of each containers in the pod before I deleted it, but for sure the pod was running and there was some problem with it.
Deleting and recreating some APIs (deployment, service, ingress and kongingress) without any change in yaml files, the output kubectl get ingress showed me these APIs without address.
At that point I deleted manually the pod ingress-kong and after it restarted, the APIs recovered the address.
I hope I was clear enough.
In the output kubectl describe for ingress-kong pod, what does it mean
Last State: Terminated
Reason: Error
Exit Code: 255
Started: Tue, 21 Jan 2020 17:17:57 +0100
Finished: Tue, 21 Jan 2020 17:17:58 +0100
Ready: True
Restart Count: 1
for ingress-controller container?
I’m testing now the same procedure and at the moment I have the same issue. kubectl logs -f ingress-kong shows me this for my new API :
W0124 14:40:10.079981 1 parser.go:772] service ********* does not have any active endpoints